Skip to content

Latest commit

 

History

History
120 lines (99 loc) · 3.36 KB

ReadMe.md

File metadata and controls

120 lines (99 loc) · 3.36 KB
Raw

Inspiration

The dbt Engineers wrote community blog post on achieving realtime analytics with dbt. This project attempts to replicate what has been done by dbt engineers. The link to the blog post is posted below:

https://discourse.getdbt.com/t/how-to-create-near-real-time-models-with-just-dbt-sql/1457

Additionally, this project also attempts to compare the above approach against Snowflake Streams and Dynamic Tables.

Reference Article:

Streams and Tasks: https://quickstarts.snowflake.com/guide/getting_started_with_streams_and_tasks/index.html

Dynamic Tables: https://quickstarts.snowflake.com/guide/getting_started_with_dynamic_tables/index.html

Approaches to real time analytics

  1. Lambda View (Using SQL Server. But it should be Database Agnostic)
  2. Snowflake Streams
  3. Dynamic Table in Snowflake (with dbt https://docs.getdbt.com/reference/resource-configs/snowflake-configs)

Steps:

Initial Setup - Directory, Python

  1. Create directory ./realtime-dbt
  2. Create virtual environment for Python python -m venv env
  3. Install dbt and dbt adapter for sql server python -m pip install dbt-sqlserver
  4. Install pyodbc and faker library
  5. pip install faker pyodbc

Git Setup

git init
git add ReadMe.md
git commit -m "first commit"
git branch -M main
git remote add origin https://github.com/aj-22/realtime-dbt.git
git push -u origin main

Database Setup

  1. Download SSMS and SQL Server Developer Edition 2022
  2. Ensure installation in Integrated Mode or change the security settings later to enable "SQL Server and WIndows Authentication Mode"
  3. Run the following query to get port number and ip address 
    USE MASTER
GO
xp_readerrorlog 0, 1, N'Server is listening on'
GO
    By default the port number is 1433. But since I have multiple instances running, my port number is 1435
  4. Run this query to create database 
    CREATE DATABASE analytics

Initialize and set-up dbt

  1. Initialize: dbt init
  2. Get location of profiles.yml: dbt debug --config-dir
  3. Enter connection parameters
analytics:
  target: dev
  outputs:
    dev:
      type: sqlserver
      driver: ODBC Driver 18 for SQL Server
      server: localhost
      port: 1435
      database: analytics
      schema: dev
      user: sa
      password: ********
      encrypt: false
      trust_cert: false
  1. Test if connection works: dbt debug

Project Set Up

  1. Update database_connect/connection.py file
  2. Execute dbt seed to load source data and dbt run to create incremental models and fresh view
  3. Run loader.py to set up event based data loading
  4. Run generator.py to create the transaction files
  5. Verify if data is getting loaded successfully in database
  6. Execute dbt run to start a batch job

Set up Snowflake and Dbeaver

  1. Create Snowflake free trial account
  2. Install DBeaver
  3. Install Git plugin for DBeaver
  4. Clone this repo in local (if not already done)
  5. Open the project from DBeaver's Git plugin
  6. Set up Snowflake Connection in DBeaver

Execution

Open three CMD Terminals

On Terminal 1, run following commands

env\Scripts\activate.bat
cd realtime-dbt\data
python loader.py

On Terminal 2, run following commands

env\Scripts\activate.bat
cd realtime-dbt\data
python generator.py

On Terminal 3, run following commands

env\Scripts\activate.bat
cd realtime-dbt\analytics
dbt run